1877.] 
of Eastern Turlcistdn. 
253 
in theory that o£ the Arabic alphabet; but as this system is a foreign adapt¬ 
ation, and not a home-growth (for the Eastern Turki existed both in a 
spoken and in a written form before the introduction of Arabic into the 
country), we find many traces of imperfect fitting between the language 
and the writing that conveys it. 
Even in books we meet with the same word spelt in different manners ; 
but more especially is the divergence apparent in the want of distinction 
between the sounds represented by the short vowel marks “ zabar” ( — ), 
“zer” ( — ), and “ pesh” (—) , and by the long vowels “ alif” (| ), 
✓ 
“ ye” (ci 1 )? and “ wao” (j ) respectively. These latter are often pro- 
nounced short, e. g. : pronounced din (from) as though written cm, 
while in Persian it would be pronounced din (faith). The “pesh” over 
a consonant is often pronounced like an i: e. g. } tush, pronounced tish 
(tooth). 
The Yarkandis on the whole seem to use the oriental vowels much 
like we use our own. That is, they rather ignore the system of vowel- 
sounds as inherent in consonants, and consider a written vowel almost 
necessary to complete a syllable, whether long or short.* 
CHAPTER II. 
PHONETIC YARIATION. 
The Root of a Yerb never changes in any formation ; nor does any 
Substantive! or Adjective. But there are numerous syllables capable of 
being affixed, whose vowels and consonants vary with those of the word 
they are attached to, according to certain general rules which are not with¬ 
out exceptions. 
A somewhat similar fact in European languages (as regards the con¬ 
sonants) is the change of the Latin prefix ad into ae in the word accept, 
into at in the word attain, and into af in the word affix, &c. 
* I have found spelt &c. 
f Excepting when a final q or k softens into a gh or g respectively before an 
affix. See p. 256. 
